1. Field of the invention
The invention relates to the access of files in electronic devices. Particularly, the invention relates to logical file names in a computer system comprising a primary memory and at least two secondary memories.
2. Description of the Related Art
User friendliness is an important factor for the success of any operating system. Especially this is true when it comes to the way files are organized and referred to in the operating system. It must be easy as well for an inexperienced user to find his files and for an experienced user to be able to quickly manage complicated directory trees.
Presently, there are two different paradigms in the organizing of file systems. The most familiar paradigm originates from the Microsoft MS-DOS™ operating system and has been evolved from there to the most recent Microsoft WINDOWS™ operating system versions. Some aspects of the Microsoft WINDOWS™ NT architecture originate from the Digital VAX/VMS™ operating system. In Microsoft WINDOWS™ separate file systems are referred to using letter symbols such as “A:”, “B:”, “C:”, “D:” and “E:”. The letter “A:” mostly refers to a disk drive for removable diskettes, “C:” to the local hard disk drive, “D:” to CD-ROM station and letter from “E:” onwards to network file systems or secondary partitions on the local hard disk drive.
The second somewhat different paradigm originates from the UNIX™ operating system. Recently the open source LINUX™ has become the most popular operating system in the family of UNIX™ look-alike operating systems. In LINUX™ operating system all files accessible within a computer are organized as a vast directory tree. Individual file systems such as diskette drives, hard disk partitions and network file systems are mounted to a given node within the directory tree. The node is called a mount point. The file systems may be mounted to an arbitrary node. Due to this fact the actual physical location of individual files is effectually hidden from the users. This may lead to various sorts of problems, for example, confidential files may accidentally be copied on a network file system accessible to other users. Similarly, it may be difficult for the user to know what files are located on the hard disk drive and what in a removable memory medium.
Reference is now made to FIG. 1, which illustrates an example of a directory tree in a LINUX™ computer. The directory tree comprises a number of subdirectories. The subdirectories are located either on a partition on a hard disk drive 100 or on a Network File System (NFS) directory 105, which is accessed via network 140. Network 140 is, for example, a Internet Protocol (IP) network. In FIG. 1 hard disk drive 100 has been partitioned to partitions 101, 102, 103 and 104. Partition 104 has been designated as a boot partition, from which the operating system kernel is loaded. Partition 104 is also assigned as the root partition, on which the root node 120 of the directory tree is located. The file system on the root partition is mounted to root node 120 during the booting process as illustrated with line 134. The file system on partition 101 is mounted to node 122, which represents subdirectory “/usr”, as illustrated with line 131. The file systems on partitions 102 and 103 are mounted to nodes 125 and 126, respectively, as illustrated with lines 132 and 133. NFS directory 105 is mounted to node 121, as illustrated with line 135. The subdirectories 123 and 124 are located on the file system on root partition 104 even though the subdirectories, in other words nodes, 125 and 126 act as mount points for the file systems on partitions 102 and 103, respectively.
The problems associated with LINUX™ file system organization as a single directory tree are evident from FIG. 1. For an inexperienced user the directory tree structure may not be very illustrative as to where files are physically located. Especially, the location of mount points 125 and 126 within the root file system may lead to confusions. Similarly, the remote nature of NFS directory 105 is not explicit for the user. However, despite the problems associated with file naming, the LINUX™ operating system has some clear advantages. Therefore, a file system would be preferable, which incorporates both the explicit drive naming from MS-DOS™ and the features of a LINUX™ file systems. Such file system should be operable within the LINUX™ operating system environment without necessitating significant changes to it. The file system and the file access therein should employ the operating system services directly from the LINUX™ operating system.